Amplify Labs
Ryno Delivery logo

Food & grocery delivery · Ghana

Making a delivery platform ready for the scale it was earning.

Ryno Delivery

The problem

Ryno Delivery runs four separate backends: customer, picker, rider and admin. The platform was growing faster than the infrastructure underneath it. AWS costs were eating margin, nobody could point to what was driving them, and the Laravel monolith was slowing feature delivery. They brought us in as technical consultants on scalability, across both the application and the infrastructure.

What we built

  • Full AWS cost audit with quantified reduction plan
  • RDS right-sizing
  • Removal of unused ElastiCache provisioning
  • Rewrite of a scheduled job firing every 30 seconds
  • Laravel to NestJS migration
  • MySQL to PostgreSQL migration

Architecture notes

The audit found the bill was structural rather than incidental: an over-provisioned database sized for a peak that never arrived, a cache cluster running unconnected, and a driver-assignment cron executing every thirty seconds regardless of whether there were orders to assign. The migration path was designed to run alongside production rather than as a cutover. Services move one at a time, with both stacks live during transition.

Outcome

‹N›% reduction in monthly infrastructure spend. Migration in progress with no customer-facing downtime to date.

Got a system like this in mind?

Tell us the operations problem behind it. Most of these projects started as a conversation about a spreadsheet, not a software brief.

Or write to us through the contact form.